ORNL/TM-2017/165

Future Muon Source Possibilities at the SNS

Gregory J. MacDougall Travis J. Williams

Date: March 28, 2017

DOCUMENT AVAILABILITY

Reports produced after January 1, 1996, are generally available free via US Department of Energy (DOE) SciTech Connect. Website http://www.osti.gov/scitech/ Reports produced before January 1, 1996, may be purchased by members of the public from the following source: National Technical Information Service 5285 Port Royal Road Springfield, VA 22161 Telephone 703-605-6000 (1-800-553-6847) TDD 703-487-4639 Fax 703-605-6900 E-mail info@ntis.gov Website http://www.ntis.gov/help/ordermethods.aspx Reports are available to DOE employees, DOE contractors, Energy Technology Data Exchange representatives, and International Nuclear Information System representatives from the following source: Office of Scientific and Technical Information PO Box 62 Oak Ridge, TN 37831 Telephone 865-576-8401 Fax 865-576-5728 E-mail reports@osti.gov Website http://www.osti.gov/contact.html

This report was prepared as an account of work sponsored by an agency of the United States Government. Neither the United States Government nor any agency thereof, nor any of their employees, makes any warranty, express or implied, or assumes any legal liability or responsibility for the accuracy, completeness, or usefulness of any information, apparatus, product, or process disclosed, or represents that its use would not infringe privately owned rights. Reference herein to any specific commercial product, process, or service by trade name, trademark, manufacturer, or otherwise, does not necessarily constitute or imply its endorsement, recommendation, or favoring by the United States Government or any agency thereof. The views and opinions of authors expressed herein do not necessarily state or reflect those of the United States Government or any agency thereof.

Approved for public release. Distribution is unlimited.

ORNL/TM-2017/165

Neutron Sciences Directorate

Report from the

Workshop on Future Muon Source Possibilities at the SNS

September 1-2, 2016

Gregory J. MacDougall

Travis J. Williams

Date Published:

March 28, 2017

Prepared by

OAK RIDGE NATIONAL LABORATORY

Oak Ridge, Tennessee 37831-6283

managed by

UT-BATTELLE, LLC

for the

US DEPARTMENT OF ENERGY

under contract DE-AC05-00OR22725

iv

CONTENTS

Page

LIST OF FIGURES ...................................................................................................................................... v LIST OF TABLES ....................................................................................................................................... vi ACRONYMS .............................................................................................................................................. vii EXECUTIVE SUMMARY .......................................................................................................................... 1 INTRODUCTION ........................................................................................................................................ 4 1. PRESENTATION SUMMARIES ........................................................................................................ 6

1.1 Alan Tennant, ORNL – ‘Neutron Sciences at Oak Ridge National Laboratory.’ ....................... 6 1.2 John Galambos, ORNL – ‘SNS: The Second Target Station (STS) Upgrade Project

Design Status.’ ............................................................................................................................ 6 1.3 Mike Plum, ORNL – ‘SNS Muon Target Possibilities.’ ............................................................. 7 1.4 Adrian Hillier, ISIS, Rutherford Appleton Laboratory – ‘Muons at ISIS.’ ................................ 8 1.5 Koichiro Shimomura, Japan Proton Accelerator Research Complex – ‘Status of J-PARC

Muon Science Experiment.’ ........................................................................................................ 9 1.6 Elvezio Morenzoni, Paul Scherrer Institute – ‘Muons for Solid State Research at the

Paul Scherrer Institute.’ ............................................................................................................ 10 1.7 Syd Kreitzman, TRIUMF Meson Facility – ‘Overview of the TRIUMF Centre for

Molecular and Materials Science.’............................................................................................ 11 1.8 Andrew MacFarlane, University of British Columbia – ‘β-detected NMR at TRIUMF:

Status and Progress.’ ................................................................................................................. 13 1.9 Yasutomo Uemura, Columbia University – ‘First-Order Magnetic Quantum Phase

Transitions in Mott Insulators and Unconventional Superconductors.’ .................................... 14 1.10 Jeff Sonier, Simon Frasier University – ‘Science with µSR: Hard Condensed Matter.’ .......... 15 1.11 Michael Graf, Boston College – ‘Magnetism in α-NaxMnO2.’ ............................................... 17 1.12 Graeme Luke, McMaster University – ‘Muon Spin Rotation/ Relaxation as a Probe of

Unconventional Superconductivity.’ ........................................................................................ 17 1.13 Stephen Blundell, Oxford University – ‘Using Muons to Learn about Fe-based

Superconductors and Novel Magnets.’ ..................................................................................... 18 1.14 Iain McKenzie, TRIUMF Meson Facility – ‘µSR and β-NMR of Soft Matter and

Chemical Systems.’ ................................................................................................................... 19 1.15 Ross Carroll, Arkansas State University– ‘µSR & Semiconductors.’ ...................................... 21

2. SUMMARY OF BREAKOUT SESSIONS ....................................................................................... 22 2.1 Source Possibilities ................................................................................................................... 22 2.2 Targets and Beamlines .............................................................................................................. 23 2.3 Science Drivers ......................................................................................................................... 24

CONCLUSIONS......................................................................................................................................... 27 REFERENCES ........................................................................................................................................... 28 APPENDIX A. Source Calculation ........................................................................................................... 29 APPENDIX B. PARTICIPANTS, PRESENTATIONS AND AGENDA ................................................. 34 APPENDIX C. COMPARISON OF EXISTING MUON SOURCES WORLDWIDE ............................ 39 APPENDIX D. INVITATION lETTER TO EXTERNAL PARTICIPANTS ........................................... 40

v

LIST OF FIGURES

Figure Page

Figure 1. A schematic of the placement of the proposed SEE facility (see Ref. 1). ..................................... 8 Figure 2. The reported distribution of science topics explored in proposals submitted to the

ISIS/RAL muon source. ............................................................................................................ 9 Figure 3. One idea suggested by Dr. Hillier for a flexible "super-µSR" beamline which could

increase count rates and potentially also timing resolution from the current generation

of µSR instruments. ................................................................................................................... 9 Figure 4. (left) The user numbers at the LMU in PSI over the last ten years. (right) The geographic

breakdown of users, using 2014 as an example year. .............................................................. 11 Figure 5. Schematic of the new detector apparatus being employed in the M9A spectrometer, which

mounts SiPM's on the scintillators in the magnet and thus eliminates the need for light-

guides. ..................................................................................................................................... 13 Figure 6. (left) The asymmetric weak decay of

8Li

+ allows this isotope to be used analogously to µ

+

to comment on magnetic properties of materials. (right) Varying the isotope kinetic

energy allows a user to choose implantation depth, and student material properties as a

function of distance from a surface or interface. ..................................................................... 14 Figure 7. (left) The normalized Bragg peak intensity as a function of pressure in MnSi. (right) The

pressure-dependence of the internal field (frequency) and magnetic volume fraction as

measured by µSR. ................................................................................................................... 14 Figure 8. A picture and schematic of the MULTI sample holder for performing μSR measurements

on multiple crystals simultaneously. ....................................................................................... 16 Figure 9. (left) Resistivity of SmB6 showing the onset of the bulk insulating state below 10 K.

(center) The temperature-dependent relaxation rate in pure and Fe-doped SmB6. The

phase transition in the parent compound is smoothed out in the Fe-doped case. (right)

The μSR spectrum as a function of field, demonstrating the existence of slow magnetic

fluctuations. ............................................................................................................................. 16 Figure 10. (left) The structure of α-NaMnO2, showing the muon stopping sites. (center0 the triclinic

distortions in the monoclinc bulk structure. (right) The field distribution based on the

calculated stopping sites. ......................................................................................................... 17 Figure 11. The magnetic field distribution in V3Si at H = 3 T (left) and H = 5 T (right). .......................... 18 Figure 12. DFT calculations quantify the effect of a µ

+ charge on the local crystal field environment

in Pr2Sn2O7 (left). Similar analysis allows for quantitative analysis of µSR data from on

a number of 227 pyrochore systems (right). Adapted from [4]. .............................................. 19 Figure 13. Example µSR data on a particular cholesterolic liquid crystal. (left) A schematic of the

cholesterol molecule, showing two muon bonding sites. (center) Time-integrated µSR

showing two distinct nuclear resonances. (right) Temperature dependence of this data

reveals the presence of two distinct phases: and isotropic phase (I) at high temperatures,

and a “narrowing phase” (N) where the molecules wobble around a specific axis thus

reducing the width of the resonance line. Adapted from [5]. .................................................. 20 Figure 14. A recent rf-field resonant µSR spectrum from which authors inferred the existence of

two shallow muonium acceptor sites in an alloy of silicon and germanium. From [6]. .......... 21 Figure 15. (left) The intensity of the proposed laser pulse as a function of time, with 30 ns at peak

intensity. (right) The resulting time structure of the muon pulse, resulting from the

convolution of the laser pulse shape with the pion lifetime. ................................................... 31

vi

LIST OF TABLES

Table Page

Table 1. The operational parameters of the SNS accelerator after the PPU project. ................................. 30 Table 2. Name, location and facility type of the 4 muon sources around the world. ................................. 33 Table 3. Table of workshop participants. ................................................................................................... 35 Table 4. Table of workshop presentations. ................................................................................................ 36

vii

ACRONYMS

BES Basic Energy Sciences

CD-1 Conceptual Design stage 1

DOE Department of Energy

ESS European Spallation Source

FAA Federal Aviation Administration

FTS First Target Station

GeV Giga electron volt (109 eV)

H-/H

0/H

+ Hydrogen with 2/1/0 electrons

HFIR High Flux Isotope Reactor

Hz Hertz (s-1

)

ISIS Rutherford Appleton Laboratory

J-PARC Japan Proton Accelerator Research Complex

Linac Linear Accelerator

mA milli Amp (10-3

A)

MeV Mega electron volt (106 eV)

MHz Megahertz (106 Hz)

MW Mega Watt (106 W)

ns nanosecond (10-9

s)

NScD Neutron Sciences Directorate

ORNL Oak Ridge National Laboratory

PPU Proton Power Upgrade

PSI Paul Scherrer Institute

QCMD Quantum Condensed Matter Division

SEE Single Event Effects

SNS Spallation Neutron Source

STS Second Target Station

THz Terahertz (109 Hz)

TRIUMF TRIUMF Meson Facility

US United States

β-NMR β-detected Nuclear Magnetic Resonance

µs Microsecond (10-6

s)

µSR Muon Spin Relaxation/Rotation/Resonance

viii

1

EXECUTIVE SUMMARY

The workshop “Future Muon Source Possibilities at the SNS” was held September 1-2, 2016 at Oak

Ridge National Laboratory. The workshop aimed to examine the technical feasibility and scientific need

to construct a µSR and/or NMR facility at the SNS. During the course of the workshop it became

evident that recently developed technology could enable the development of a world leading pulsed muon

source at SNS, without impacting the neutron science missions of the SNS. The details are discussed

below.

Muon Spin Rotation/Relaxation/Resonance (µSR) is a technique that involves using spin-polarized

muons implanted in a material to provide extremely sensitive measurements of the static and dynamic

properties of the local magnetic field distribution within the sample. detected Nuclear Magnetic

Resonance (NMR) is a related technique that employs spin-polarized 8Li ions in the place of muons.

These techniques have led to important results in condensed matter physics, chemistry and semiconductor

physics, among other fields. In the study of magnetism especially, this technique is highly complementary

to neutron scattering: where neutrons provide bulk measurements in reciprocal space with fluctuations on

the timescale of nanoseconds (THz), µSR is a local, real space probe, sensitive to microsecond-scale

fluctuations (MHz). The two techniques share a common user base, and for this reason 3 of the 4 existing

µSR facilities in the world are co-located with neutron sources (PSI, ISIS and J-PARC). The glaring

exception is in North America, where the sole muon source is located at the meson accelerator laboratory

TRIUMF in Vancouver, Canada. The United States has never developed a similar user facility. To

address this national shortcoming, there have been several efforts in recent years to assess the feasibility

of building a US facility for µSR, including a conversation during the construction of the first target

station (FTS) of the SNS (2000), and more recently in the context of ProjectX at FermiLab (2013) and the

Transformative Hadron Beamlines initiative at Brookhaven National Laboratory (2014). At ORNL, recent

successes at the SNS, along with the ongoing PPU and STS expansion projects, have motivated a larger

conversation about future science possibilities at the lab. These conditions and the clear national interest

have motivated the current workshop to reassess the feasibility of a µSR source co-located with the SNS.

The workshop brought together 15 invited researchers from universities and existing SR facilities,

along with 8 ORNL staff with detailed knowledge of the SNS accelerator systems, the existing neutron

program, and ongoing discussion of the PPU and STS projects. The goals of the workshop were to

discuss the scientific challenges that can be addressed by neutrons together with a muon / ion source, the

technical requirements for building such a facility, and the perspective of the μSR user community on

their desire for such a source. The discussions during the workshop were productive, touching on several

topics including the state of the μSR community, high-impact science, complementarity with neutron

scattering, and muon source possibilities. Particularly promising were discussions surrounding emergent

plans for a unique and world leading pulsed muon source at the SNS that could co-exist with the neutron

program. Main conclusions of the workshop were as follows:

μSR is a mature probe of condensed matter which continues to produce high-impact

science. Measurements using the μSR technique have been able to solve outstanding problems in

quantum magnetism, chemistry, superconductivity, and semiconductor physics. In studies of

magnetism, the sensitivity of μSR to minute magnetic fields is leveraged to measure ordered

moments as small as 10-3

μB and emergent local fields associated with novel states breaking time-

reversal symmetry. Furthermore, the local nature of μSR allows it to be used to detect short

range magnetic correlations and to directly determine ordered volume fractions- a strongly

complementary measurement to neutron scattering. In superconductors, μSR has been shown to

2

be an extremely sensitive measurement of magnetic penetration depth. This absolute measure of

the penetration depth is regularly used to calculate the superfluid density as a function of

temperature, allowing for strong statements about the nature of the superconducting gap

symmetry. For applications to chemistry, it is important to note that muons are long-lived enough

to capture an electron, forming muonium. The muonium particle then can then simulate hydrogen

and can be used to probe chemical reaction rates. The energy scales of μSR also make it well-

suited to measuring hyperfine coupling constants and level-crossing resonances in chemical

systems. In the study of semiconductors, muon hopping rate is probed, which allows for a

sensitive measure of quantum diffusion processes, as well as the acceptor/donor state properties

on the nanoscale.

It was strongly emphasized that the co-location of µSR and neutron sources is mutually

beneficial to both communities. The international workshop attendees pointed out that three of

the four active μSR facilities (ISIS, J-PARC and PSI) are co-located with neutron sources, with

which they often shared user programs. It was estimated that 10-15% of neutron users at co-

located sources also use μSR to study their materials. This complementarity has been strongly

utilized by researchers outside of the United States, but also US researchers, albeit on a less

frequent basis due to the need for international travel. Several presenters pointed to recent and

active collaborations with US neutron scatterers, indicating a growing interest in μSR from the

US neutron community. Moreover, there has been a recent increase in publications that make

use of both techniques, allowing for more impactful conclusions than either technique could have

supported alone. It is expected that co-locating muon and neutron sources in the United States

would facilitate a large expansion in the μSR user base.

There have been several recently-developed technologies and routes of development that

can be leveraged in a new μSR facility. These technologies would fit very well into the

technical portfolio of ORNL and the other DOE/BES national laboratories, and will be important

in overcoming current limitations of the μSR technique. This would position a μSR facility at

ORNL to have world-leading measurement capabilities. These include RF-techniques that can be

used to provide an order of magnitude improvement on timing resolution when measuring in high

applied magnetic fields- a particular problem in a technique like μSR that makes use of charged

particles. In the measurements of semiconductors, the ability to do in-situ optical measurements

would increase the capabilities of μSR measurements, while other in-situ techniques can be used

for pump-probe and out-of-equilibrium measurements. The development of segmented detectors

for μSR measurements has been recently proposed as a means of providing enhanced spatial

resolution in measurements and could be employed to develop new muon microscopy facilities.

Finally, as with neutron beamlines, μSR measurements have increasingly high demands for

extreme sample environments: temperature extremes, high magnetic fields and high pressures.

The existing DOE/BES system is well-positioned to address these technical challenges at a

cutting-edge μSR facility.

Several possibilities for the location and nature of the muon source at the SNS were

considered at the workshop, with one clearly superior than the others: o The first possibility was to place a muon target in the primary proton beam, prior to the

beam entering the FTS or STS buildings. This was one method that was considered

during the muon source assessment prior to the construction of the SNS, and is the

method currently used at ISIS, PSI and J-PARC. This possibility was rejected for two

reasons: its impact on neutron production and the timing structure of the proton beam.

Addressing the former point, it was shown that PSI and ISIS lose ~30% of the protons

that would go to neutron production with the muon target in the incident beam. Of this

loss, only 2% of the protons are used to produce muons, while the other ~28% are lost

3

from scattering off the target. On the latter point, the 1ms pulse is far too long for muon

production and could not be suitably modified to the desired muon timing structure

without significant losses in beam intensity. The ideal muon pulse for μSR measurements

is approximately 30ns, due to the pion lifetime.

o The second possibility that was considered was to utilize the protons that are currently

being directed to the beam dump. This has the advantage that it has no impact on neutron

production, but the beam would still have the unfavorable timing structure. The pulse

can be made useful for μSR by chopping or splitting the beam, however such beam

manipulations severely impact the flux. It was decided that the flux of a muon beam

using protons from the beam dump would not be high enough to justify the construction

of a μSR facility. However, the timing structure of a β-NMR experiment does not suffer

these limitations, and so the protons going to the beam stop could be used to create a

world-class β-NMR facility.

o The last source possibility was developed and explored at the meeting, and involves the

use of laser-stripping to take extremely short proton pulses from the beam immediately

prior to entering the accumulator ring. This idea had been previously proposed for the

SEE Systems Test Facility proposed for use by the FAA. This technology has been

previously demonstrated, and it could be used to create a proton pulse of any duration

with ~70-90% stripping efficiency. The workshop attendees refined this design and

concluded that such a method of producing a muon beam would have negligible impact

on neutron production (0.15% of the protons would be diverted), would have the best

resolution and highest flux of any pulsed muon source in the world. The existing

infrastructure of accelerator systems, personnel and technologies means that this world-

leading muon source could be constructed with substantial reductions in time, money and

technological development relative to a stand-alone facility.

4

INTRODUCTION

Muon Spin Rotation/Relaxation/Resonance (µSR) is a technique that involves using spin-polarized

muons implanted in a material to provide extremely sensitive measurements of the static and dynamic

properties of the local magnetic field distribution within the sample. Proton beams are directed into a

low-Z target (typically Carbon or Beryllium) to produce pions. The pions decay with a mean lifetime of

26 ns via the weak interaction into a muon (or antimuon) and an anti-muon neutrino (or muon neutrino).

Muons produced from pions at rest (residing on or near the surface of the target) are known as surface

muons, and are the most commonly used muons for µSR experiments. The muon carries most of the

momentum from this decay, but will stop in fairly thin samples, with a stopping range of ~120 mg/cm2 in

carbon. Conversely, muons produced from pions that have been ejected from the target are known as

decay muons. These decay muons are useful because of their higher momentum, resulting in deeper

penetration into samples and allowing for measurement of materials in an enclosure or high pressure

environment. Finally, beams of low energy muons can be produced by taking beams of surface muons

and reducing their momentum. This is currently done successfully by passing the beams through thin

sheets of noble gases at low temperatures, but there is a developing effort to slow muons using laser

pumping. Due to the sharp drop in the muon flux during the slowing process (4-5 orders of magnitude), a

high initial flux is needed to make such beamlines feasible. Beams of low-energy muons have a much

shorter stopping distance, allowing measurements of thin films, nanostructures and surface properties.

Because the pion decay is governed by the weak interaction, which violates parity, both the neutrinos

and muons produced are exclusively left-handed; that is, their spin is antiparallel to their linear

momentum. Thus, the surface muons will be 100% spin polarized, while decay muons have marginally

lower (>80%) spin polarization since they have to be separated in-flight from pions and

electrons/positrons. Depending on the charge of the pion, either positive or negative muons can be

produced and used to perform µSR experiments. However, since negative muons have more complex

interactions with the samples being measured, nearly all µSR experiments tend to be performed with

positively-charged antimuons. For the remainder of the report, “muons” could refer to either negatively-

or positively-charged muons, but for practical purposes, we strongly emphasize the use of positively-

charged muons. detected Nuclear Magnetic Resonance (NMR) is a related technique that employs

spin-polarized 8Li ions in the place of muons.

Both techniques have led to important results in condensed matter physics, chemistry and

semiconductor physics, among other fields. In the study of magnetism especially, this technique is highly

complementary to neutron scattering: where neutrons provide bulk measurements in reciprocal space with

fluctuations on the timescale of nanoseconds (THz), µSR is a local, real space probe, sensitive to

microsecond-scale fluctuations (MHz). The two techniques share a common user base, and for this reason

3 of the 4 existing µSR facilities in the world are co-located with neutron sources (PSI, ISIS and J-

PARC). The glaring exception is in North America, where the sole muon source is located at the meson

accelerator laboratory TRIUMF in Vancouver, Canada. The United States has never developed a

competitive facility. To address this national shortcoming, there have been several efforts in recent years

to assess the feasibility of building a US facility for µSR, including a conversation during the construction

of the first target station (FTS) of the SNS (2000), and more recently in the context of ProjectX at

FermiLab (2013) and the Transformative Hadron Beamlines initiative at Brookhaven National Laboratory

(2014). At ORNL, recent successes at the SNS, along with the ongoing PPU and STS expansion projects,

have motivated a larger conversation about future science possibilities at the lab. These conditions and the

clear national interest have motivated the current workshop to reassess the feasibility of a µSR source co-

located with the SNS.

5

This report summarizes the outcome of the workshop, “Future Muon Source Possibilities at the

SNS” held in the Clinch River Cabin at Oak Ridge National Laboratory (ORNL) on September 1 and 2,

2016. The workshop was designed to consider the feasibility and community interest in the construction

of a Muon Spin Rotation/Relaxation/Resonance (SR) and/or detected Nuclear Magnetic Resonance

(-NMR) facility co-located with the Spallation Neutron Source (SNS) as a complementary probe of

materials, which would enhance the overall user programs in materials research. The workshop brought

together 15 invited researchers from universities and existing SR facilities, along with 8 ORNL staff

with detailed knowledge of the SNS accelerator systems, the existing neutron program, and ongoing

discussion of the Proton Power Upgrade (PPU) and Second Target Station (STS) projects. The letter of

invitation to participants (see Appendix C), explained that the goals of the workshop were to discuss the

scientific challenges that can be addressed by neutrons together with a muon / ion source, the technical

requirements for building such a facility, and the perspective of the μSR user community on their desire

for such a source.

The scientific program and agenda was organized by G.J. MacDougall (Illinois) and T.J.

Williams (ORNL). The agenda (see Appendix B) consisted of: technical talks outlining the existing

facilities at the SNS, along with the proposed PPU and STS projects; an overview of the existing muon

facilities at the TRIUMF Meson Facility (TRIUMF, Canada), Rutherford-Appleton Laboratory (ISIS,

United Kingdom), Paul Scherrer Institute (PSI, Switzerland) and the Japan Proton Accelerator Research

Complex (J-PARC, Japan); scientific presentations on the role of µSR in condensed matter physics,

chemistry and semiconductor physics; and breakout sessions aimed at discussing different possibilities for

a putative muon source at the SNS, and necessary conditions such a source must have to address of

current and future scientific interest. This report summarizes the presentations from the workshop in

Section 1. Section 2 summarizes the conclusions from the breakout sessions, including a preliminary

design for a muon source and a comparison of the relative figures of merit to other sources around the

world. We then conclude with a summary of main outcomes from the workshop and potential paths

forward.

6

1. PRESENTATION SUMMARIES

1.1 ALAN TENNANT, ORNL – ‘NEUTRON SCIENCES AT OAK RIDGE NATIONAL

LABORATORY.’

Dr. Tennant, Chief Scientist, Neutron Scattering Sciences Directorate (NScD), provided an overview

of the neutron scattering facilities and program at ORNL. He outlined the two current neutron sources,

the High Flux Isotope Reactor (HFIR) and the Spallation Neutron Source (SNS). HFIR was initially

constructed to meet the need for production of transuranic isotopes, but its core mission today is scientific

studies that utilize neutron scattering. While still a significant source of isotope production (primarily Cf-

252) and being used for neutron irradiation and activation analysis, the 12 neutron scattering instruments

in the user program at HFIR provide the primary scientific output of the facility. This is due to the high

neutron flux of both thermal and cold neutrons. The SNS is a complementary source, being the most

powerful accelerator-based neutron source in the world. While being designed as a neutron scattering

source from its conception, it was designed to be able to access a greater dynamic range of measurements

than HFIR. It currently has 17 instruments in the user program, with 2 others in commissioning and

another instrument in the design stage.

Dr. Tennant provided several examples of the scientific productivity of the neutron scattering

program at ORNL, including measurements of Fe-based superconductors, thermoelectric materials,

materials intended for biomass/bioenergy applications, quantum magnets, energy storage materials and

thin film heterostructures. He showed that the number of publications generated by the neutron scattering

instruments is continuing to increase, as is the oversubscription rate of both facilities. This is evidence of

the vital role that neutron scattering plays in materials research, providing information that cannot be

obtained by other techniques. This ability is recognized globally, with several recent investments in new

or upgraded facilities, including a power upgrade at J-PARC and a new facility being constructed in

Europe, the European Spallation Source (ESS). Likewise, ORNL is investing in two projects: an upgrade

to the accelerator allowing operation at higher energy and current, called the Proton Power Upgrade

(PPU), and the construction of a second target building to house a planned 22 next-generation neutron

instruments, called the Second Target Station (STS).

He emphasized that the success of the current and future neutron sources relies on the continued

engagement of the scientific community and a long-term vision of how these facilities can address

national and international scientific priorities. This is vital to continuing to attract the frontier

experiments and the most positive reflection of the facilities’ performance.

1.2 JOHN GALAMBOS, ORNL – ‘SNS: THE SECOND TARGET STATION (STS)

UPGRADE PROJECT DESIGN STATUS.’

Dr. Galambos is the Director of the Spallation Neutron Source Second Target Station Project. He

provided an overview of the STS and PPU projects, which have been separated organizationally, but are

both required for the advancement of neutron science at ORNL. He described the increased power

capabilities that will be derived from the PPU project, including upgrading from 1.4 to 2.8 MW,

increasing the neutron flux at the FTS from 1.4 to 2.0 MW; the additional 800 kW will be allocated to the

STS. This, combined with cutting-edge technology in beamline and detector technology, will make the

STS the world’s highest peak brightness neutron source, and will be optimized for cold neutrons. He

cautioned that adding a muon target in front of the STS neutron target would have a large impact on the

beam power delivered to the neutron target. As an example, he cited the dual muon/neutron source at

7

PSI, where 1-2% of the protons are used to create muons, but the proton flux at the neutron target is

reduced 20-30% due to scattering from the muon target upstream.

While outlining the development plan and timeline, Dr. Galambos emphasized that there was a path

towards developing a muon source at ORNL. He particularly emphasized two points: the need for a

strong science case and its impact on the STS design and performance. The science case for STS has

been strongly put forward through a series of workshops and reports over the past several years. This has

engaged the wider user community in many different scientific fields in order to demonstrate the need for

a balanced, diverse instrument suite. The science case for a muon source must be similarly demonstrated,

in terms of its impact, breadth and productivity. With the design of the STS building and instrument

layout essentially completed, a muon source design must be made to work within those parameters. Any

impact on the design and performance of the neutron source would require strong justification, and would

be one of the primary considerations in the approval process for the muon source. Furthermore, with the

CD-1 review for the PPU project targeted for mid-2017, there is a short timeline when a muon source

design could be considered. However, the long schedule for the accelerator upgrade construction (likely

to last through 2024) provides an ideal opportunity for the modification and construction necessary to

create a muon source at the SNS.

1.3 MIKE PLUM, ORNL – ‘SNS MUON TARGET POSSIBILITIES.’

Dr. Plum, the Accelerator Physics Team Leader in the Research Accelerator at ORNL, provided an

overview of the linac and accelerator systems. He described the pulse structure and how the H- pulse

from the linac was divided between two beam stops and the injection ring. By design, the ring injection is

only 95% efficient, such that 95% of the beam goes to the accumulator ring (as protons, H+), 3% goes to

the primary beam stop (as H0) and the remaining 2% goes to the secondary beam stop (as H

-). Either of

these two beams could be diverted to a muon target. However, the pulse going to the muon target would

have the same time structure as the pulse from the linac, which is ~1 ms wide. Using the portion of the

beam going to the beam dump would allow for a muon source to be located on the opposite site of the

accumulator ring from the neutron targets. This has the advantage of not interfering with the design or

performance of the neutron program.

A proposal had been suggested in the past for such a facility, that would be used to test Single Event

Effects (SEE), proton damage to electronic components1, however it was ultimately not constructed. A

schematic of this proposed facility and its placement relative to the SNS linac and accumulator ring is

shown in Figure 1. This facility was designed to be ~10,000 sq. ft., which would easily accommodate a

muon source and beamlines instead. While discussing the example of the SEE facility, it was noted that

the facility was not designed to use protons from the beam dump, but to utilize a method of laser-stripping

the protons to obtain a very short proton pulse in the SEE beamline(s)Error! Bookmark not defined.

. This method

as estimated to be 90-95% efficient and could be used to produce a pulse that was arbitrarily short. This

method was suggested as a viable alternative to using the protons from the beam dump, and was

subsequently acknowledged as has vastly superior beam characteristics for a muon source. Both options

were compared during the later break-out sessions, and the conclusion was reached to investigate the laser

stripping as the primary method. For further discussion, see the summary of the breakout sessions in

Section 2 and the source calculation in Appendix A. 1

8

1.4 ADRIAN HILLIER, ISIS, RUTHERFORD APPLETON LABORATORY – ‘MUONS AT

ISIS.’

Dr. Hillier is the Muon Group Leader for the ISIS facility at Rutherford Appleton Laboratory (RAL)

in Oxford, UK, and was slated to present information about the state of the µSR program at RAL. Due to

last minute personal conflicts, Dr. Hillier was unable to travel to Oak Ridge and his slides were presented

to the workshop by Prof. Stephen Blundell. The slides summarized basic information about the ISIS

muon source and user base. ISIS is pulsed muon source, created by diverting 4% of the protons destined

for the spallation neutron target. Pulses have 50Hz repetition and 70ns full-width. Impact on the neutron

program is said to be minimal. There are two sides to the µSR program at ISIS. The ‘EC muon facility’,

funded by through the European Commission, employs only positive surface muons and has three

spectrometers for materials studies. The ‘RIKEN-RAL muon facility’, operated jointly with the RIKEN

lab in Japan, uses both positive and negative muons with variable momentum and operates two

spectrometers for materials studies, as well as beam for a low-energy muon development and fundamental

physics studies. The existence of a timing structure in the proton beam allows for a large increase in muon

flux, limited only by the ‘dead time’ of positron counters. This latter constraint has led to the extensive

use of segmented counters, which increases count rates by a further factor of 5. Similar efforts should be

considered for any high-flux pulsed muon source.

Dr. Hillier noted the science areas explored in the µSR program at ISIS, which broadly separates in to

studies where muons act as a passive probe (e.g. studies of magnetism, superconductivity, molecular

dynamics, charge transport) and studies where the muon is playing an active role (e.g. semiconductors,

proton conductors, light particle diffusion). He further gave an interesting breakdown of the size and

composition of the user community at RAL, which comprises 60 different groups from 15 countries in

Europe. Roughly 30% of submitted user proposals sought to study inorganic (21%) or organic (9%)

magnetism, 23% inorganic (21%) or organic (2%) superconductivity, 11% elemental analysis, 10%

Figure 1. A schematic of the placement of the proposed SEE facility (see Ref. 1).

9

batteries or solar cells, and the remainder covering a host of topics including chemistry, semiconductors,

electronic radiation and quantum effects.

The presentation concluded with a discussion of one future path forward for muon pulsed muon

facilities. It was acknowledged that the major constraint of such facilities is the timing resolution, which

is set by the inherent time width of the incident proton (and thus muon) pulses. Dr. Hillier suggested that

resolution can be improved by temporally ‘slicing’ the muon beam to effectively decrease pulse width,

and presented a plan for a “SuperMuSR” beamline at ISIS that could operate in "High Resolution" mode,

with 10× the resolution and 2× the current count rate, or "High Rate" mode, with 15-20× the current rate.

A survey of the requirements for various µSR experiments revealed that some would benefit more from

rate and some more from resolution. Thus, it was concluded that there would be a distinct benefit in

having a muon facility able to vary from one mode to the other. This possibility was discussed

extensively in the context of a potential new source at the SNS.

Figure 2. The reported distribution of science topics explored in proposals submitted to the ISIS/RAL muon source.

Figure 3. One idea suggested by Dr. Hillier for a flexible "super-µSR" beamline which could increase count rates and potentially

also timing resolution from the current generation of µSR instruments.

1.5 KOICHIRO SHIMOMURA, JAPAN PROTON ACCELERATOR RESEARCH

COMPLEX – ‘STATUS OF J-PARC MUON SCIENCE EXPERIMENT.’

Dr. Shimomura attended as a representative for the Muon Science Establishment (MUSE) in the

Material and Life Science Facility (MLF) at the Japan Proton Accelerator Research Complex (J-PARC).

He began by laying out the source at J-PARC and the instruments in MUSE. The muon program at J-

PARC is parasitic to the neutron spallation source, and uses the same 25Hz proton beam where each

timing cycle contains a pair of 100ns wide pulses are separated by 600ns (a “double bunch” structure).

Muons are produced using a rotating graphite target, which has been operating since September, 2014

with no reported problems. As designed, the facility will ultimately have four muon beamlines with

different purposes, but they are in varying stages of construction.

10

Dr. Shimomura discussed the purpose and design specifics of each of the four beams. The most detail

was given about the ‘D-line’, a multi-purpose beam where either surface (30MeV/c) or decay (5-

120MeV/c) muons can be extracted. This beamline was first commissioned in 2014, soon after the

graphite source was installed, and is now fully in the user program. Detail was given about spectrometer

designs, all of which use segmented detectors, and the newest of which is able to count 200M

events/hour. It was clear that effort has been put into developing proper sample environments, including a

recent addition of a dry dilution refrigerator which can be cooled in 1.5 days without need for human

intervention. Dr. Shimomura then discussed the ‘U-line’, a dedicated ‘ultra-slow muon’ beam with high

intensity and high luminosity. This beamline has 10× the incident flux as the D-line, and muons are

slowed to final momenta in the range 0.05-50 keV/c using laser pumping techniques. The U-line has long

been in development, but current milestones have been hit in the past year, and as of 2016, the beam is

capable of providing a usable rate of 40 ultra-slow muons/second. Also discussed were the ‘S-line’, a

second surface muon beam that began commissioning in October, 2015, and the ‘H-line’, which is in

planning stages but is to include instruments for muon g-2 experiments, muon transmission microscope

and Mu-HF measurements. Despite the facility being in development stages, Dr. Shimomura gave

encouraging user numbers. He stated that the one instrument in the user program, D1, had 108.5 days of

operation in 2015 and received 40 proposals for that time. The instrument was a factor of 2 over-

subscribed. The user base was 90% Japanese, with the other 10% coming mostly from China and Korea,

demonstrating a strong role for geographic proximity.

On a final note, Dr. Shimomura mentioned that J-PARC has recently undergone a power upgrade to

1 MW, and is in the planning stages for a second neutron/muon target station. The muon program is

expected to benefit greatly from this development, with projected muon intensities a factor of 50 greater

than the first target station (factor of 10 from W target, and 5 from beamline capture). The large increase

in flux is leading to consideration of new classes of muon instrumentation, including an ultra-high-

intensity muon beam stroboscope for muon imaging experiments.

1.6 ELVEZIO MORENZONI, PAUL SCHERRER INSTITUTE – ‘MUONS FOR SOLID

STATE RESEARCH AT THE PAUL SCHERRER INSTITUTE.’

Dr. Morenzoni presented the state of the program at the Laboratory for Muon Spin Spectroscopy

(LMU) at the Paul Scherrer (PSI) Institut in Villigen, Switzerland. His talk consisted of an overview of

the instruments in the LMU program, a summary of scientific output, and select science examples. At

PSI, muons are created by diverting a fraction of protons from the spallation neutron target (0.4GeV,

2.4 mA = 1.5×1016

protons/second) towards a rotating graphite target to create quasi-continuous beams of

surface (4MeV), decay (10-60MeV), and ultra-low energy (1-30keV) muons. Surface muons are 100%

polarized, and these channels have 107-10

9 muons/second for condensed matter studies. Decay muons are

80% polarized, and energies can be chosen in the MeV range for depth-ranging muons into samples

contained in pressure cells or other enclosed environments. Ultra-low energy muons are used for depth-

dependent studies in the range 2-300 nm from a sample surface, and LMU supplies 1.9×108 µ

+/s towards

a dedicate low-energy beamline for this purpose.

Dr. Morenzoni describes in detail the six instrument end stations in the LMU program. Each end

station has a dedicated spectrometer, and as with other facilities, there was a strong emphasis on the

importance of sample environment capabilities. There are four instruments for utilizing surface muons:

HTF, a new spectrometer capable of measuring in fields as high as 9T while applying temperatures below

20mK; LTF, a spectrometer dedicated to measurements at dilution refrigerator temperatures in fields

below 0.6T; and GPS and DOLLY, two general purpose instruments for performing measurements in

fields below 0.6T and temperatures above 250mK. GPS and LTF share a single beamline, where an

11

electrostatic kicker only sends a muon to a specific instrument when needed (Muons On Request), thus

lowering background to levels more commonly associated with pulsed muon sources. LMU also has a

dedicated decay muon channel, GPD, used primarily for studies in pressure cells (max pressure of

2.8GPa). A series of impressive examples demonstrated how these capabilities have allowed for impactful

science with µSR at PSI. This included determination of phase diagrams and demonstration of co-existing

superconducting and magnetic volumes in FeAs materials, detection of 70G time-reversal symmetry

breaking fields in the non-centrosymmetric superconductor SrPtAs, contribution to discovery of a spin-

liquid phase in Ca10Cr7O28, and studies of the quantum critical point in quasi-skutteridite R3T4X13 with

T={Ir,Rh}. Examples concentrated on magnetism and superconductivity, reflecting Dr. Morenzoni’s own

training, but he made it clear there was also interesting work being performed at LMU in the areas of

chemistry, soft-matter, applied physics and semiconductor physics, among other areas.

A significant amount of time was spent discussing the dedicated low-energy µSR (LE- µSR )

beamline, which for several years has been performing novel and cutting-edge studies of thin-film

samples and heterostructures. The LE- µSR program at PSI is unique in the world, and uses 1.9×108 µ

+/s

on a dedicated surface muon source in conjunction with a series of moderators, kickers and electrostatic

fields to create a beam of muons with kinetic energies 0.5-30keV at a typical rate of 4500 µ+/s at the

sample site. By varying kinetic energy, one can vary the depth from the sample surface a muon will

implant. Dr. Morenzoni presented interesting applications of these new techniques, including the

measurement of spin diffusion lengths in organic spin valves, observation of a giant proximity effect in

LaCu2O4 multilayers, and measurement of an intrinsic paramagnetic Meissner effects in an odd-

frequency superconductors. As the newest item in the µSR toolbox, he made clear that it was opening up

the technique to new scientific communities.

Overall, a positive picture was painted of state of the user community in the LMU at PSI. Over the

past decade, the number of user proposals received at the institution has increased 70%, and in 2015 they

received 207 proposals to cover 763 total experiment days. Instruments were all oversubscribed, by

factors ranging from 2.0-3.5. About 60 papers are published each year containing data from experiments

at LMU, including 10 in journals that are traditionally labelled as “high impact”. The user community is

50% based in Switzerland and 30% elsewhere in Europe, again reinforcing the huge positive impact of

geographic proximity, though Dr. Morenzoni noted that about 5% of the user base is travelling each year

from the United States.

1.7 SYD KREITZMAN, TRIUMF MESON FACILITY – ‘OVERVIEW OF THE TRIUMF

CENTRE FOR MOLECULAR AND MATERIALS SCIENCE.’

Dr. Kreitzman is the Manager of the Centre for Molecular and Materials Science (CMMS) at the

TRIUMF meson facility in Vancouver, Canada, which is the sole µSR user program in North America.

His talk focused on programmatic elements of the µSR and β-NMR facilities at TRIUMF, giving detailed

Figure 4. (left) The user numbers at the LMU in PSI over the last ten years. (right) The geographic breakdown of users, using

2014 as an example year.

12

information about the number of staff, beamlines, spectrometers, and the state of various repairs and

developments. TRIUMF has 6 scientific staff and 4 technical staff members, funded largely via 5 year

contracts. It currently has two operational beamlines in the µSR program, each of which supplies a

continuous beam of surface muons. The M15 beamline began operation in 1984 and is one of the most

productive in the program, though it is beginning to show problems related to its age, reduced rates due to

the degradation of permanent magnets in the beamline optics. The other beamline, M20, was recently split

into two separate end-stations with a high-frequency “kicker” providing muons to both. The facility also

has a third beamline, M9, split into surface muon (M9A) and decay muon (M9B) channels. The decay

channel has historically been used for chemistry and high-pressure experiments, but unfortunately it has

been out of commission for several years due to a fundamental issue with the alignment of the associated

muon target. In addition to the muon facility, the CMMS has in recent years been maintaining a β-NMR

program, which uses particles from the rare isotope facility at TRIUMF to explore thin-films and surface

physics. Due to beam availability, the β-NMR program only operates 5 weeks/year.

The CMMS program at TRIUMF is undergoing an unprecedented period of expansion, and Dr.

Kreitzman detailed several ongoing and future developments. In addition to the recent doubling of

capacity on M20 from splitting the beam, a Muons on Request setup, is being commissioned and should

lower the overall background. On M15, all of the magnet power supplies and heat sinks have been

replaced. Detail was given about the newly constructed helium liquefier, which is expected to greatly

reduce user costs. One of the two M9 end-stations (M9A) is due to come online in 2017 and should direct

a new channel of surface muons towards a versatile 3T spectrometer. In addition, a consortium of

university-based research professors have recently submitted a $10M proposal to the Canadian

Foundation for Innovation to upgrade and expand the decay channel (M9B), and also provide a reliable

source of negative mons for fundamental physics and chemistry applications. With the opening of the new

ARIEL isotope facility in 2020, the β-NMR beamlines are set to triple the available beamtime and greatly

expand the program.

The functionality of the six different spectrometers in the program was detailed and, as elsewhere,

particular emphasis was placed on sample environment capability. Dr. Kreitzman discussed a new high-

magnet field spectrometer, NuTime, which has just entered the program and is capable of measuring in

fields as high as H=7T with 1.5ppm field homogeneity at the sample site. He highlighted the dilution

refrigerator, which is capable of cold sample changes, can apply fields of up to 3T, and has base

temperature of 12mK. Coming up, he spoke about a new generation of spectrometers which employ

active muon collimation and scintillator-mounted Si photo-multipliers, eliminating light guides and

allowing for lower background and timing resolution as small as 50ps. A detailed plan was laid out for

new β-NMR spectrometers and facilities, including the development of a liquid Mg isotope target argued

to be ideal for biological studies. Finally, Dr. Kreitzman detailed the extensive outreach and community

expansion efforts being supported by the CMMS program, which aims to raise awareness of µSR in the

U.S. institutions and broaden in the North American user community in coming years.

Throughout his talk, Dr. Kreitzman spoke positively of the prospect of a muon source of SNS, and

specifically pointed to the strong collaborations between µSR and American neutron scattering user

bases. He mentioned that there are 48 separate U.S. neutron scatters that have been users at TRIUMF in

recent years. He also spoke of the complementarity between the continuous and pulsed muon facilities,

and suggested such complementarity would extend to the source at TRIUMF and any pulsed facility at

ORNL. The SNS, he pointed out, naturally has the timing characteristics and expertise in RF technology

to build a world-class Muon Magnetic Resonance program, akin to the program at ISIS, and could build a

pulsed facility that would put ORNL on a “Pinnacle of the MuSR World”.

13

Figure 5. Schematic of the new detector apparatus being employed in the M9A spectrometer, which mounts SiPM's on the

scintillators in the magnet and thus eliminates the need for light-guides.

1.8 ANDREW MACFARLANE, UNIVERSITY OF BRITISH COLUMBIA – ‘β-DETECTED

NMR AT TRIUMF: STATUS AND PROGRESS.’

Dr. MacFarlane, Associate Professor in the Department of Chemistry at the University of British

Columbia introduced the technique β-detected NMR (β-NMR), and discussed its scientific impact and

future directions for the technique. The β-NMR program at TRIUMF is unique in the world, and uses the

rare-isotope production facility there to create spin polarized beams of radioactive isotopes which are then

implanted in materials of interest, evolve in the local field environment and undergo β-decay. The

technique has strong parallels with µSR, and obtains similar but distinct pieces of information. The longer

decay times of isotopes make them sensitive to fluctuations on the timescale of seconds, rather than

microseconds for muons, making them better suited for measurements of spin-lattice relaxations in

metals, for example. Further, the low kinetic energies involved allow for controlled implantation depths

of isotopes over the range 100-2000Ǻ, making the technique a natural complement to neutron

reflectometry in the study of surface or interface magnetism. Dr. MacFarlane talked about isotope

production at TRIUMF, which involves irradiating specially design tantalum targets at temperatures

exceeding 2000C with 500MeV protons. This process produces high quality beams of isotopes with

diameter 4-8mm at a rate of ~107 ions/second. Ion beams are then actively spin polarized using circularly

polarized laser light, and momentum is chosen by maintaining an electric potential difference between the

ion target and sample. The most commonly used isotope is 8Li

+, which is sensitive to electric quadrupoles

in addition to spins, but researchers at TRIUMF also have access to spin-only 31

Mg2+

ions. Typical

measurements last ~1hr, comparable to µSR.

Though now fully in the user program at TRIUMF, β-NMR currently runs on a restricted schedule of

5 weeks of beamtime per year. Dr. MacFarlane emphasized that the technique is still in its early days,

comparable to the state of µSR in the early 1980s, and that much of the early research has focused on

finding best operating procedures and basic characterization of spectrometers properties such as

backscattering and depth profiles. He proceeded, however, to give examples of important scientific results

from recent years. This included measurements of surface effects in Fe2O3 below the Morin transition,

second timescale spin fluctuations in correlated metal Sr2RuO4, Li+ diffusion at ionic heterointerfaces,

and orbital/spin response in topological insulator Bi0.9Sb0.1. The newly constructed ARIEL isotope facility

at TRIUMF is slated to contain a dedicated β-NMR beamline.

Dr. MacFarlane communicated that lessons learned at TRIUMF could guide the design and

construction of a similar β-NMR facility at the SNS. For example, he notes that the Ta-foils used to

produce 8Li

+-ions have a 6-week lifetime, and specially designed targets and protocols would need to be

developed to change “hot” targets on a semi-regular basis. Tuning and beamline alignment, currently a

14

limiting step at TRIUMF, would be helped by modern technologies. He further emphasized that the main

source of errors limiting data are systematic in nature, not statistical, and that beam stability should be

emphasized over raw flux in any future design. The 60Hz repetition rate and pulse characteristics in the

FTS or beam dump which are quite limiting for the design of a muon source, are considered suitable for

β-NMR, to which a 60Hz timing structure would appear quasi-continuous.

1.9 YASUTOMO UEMURA, COLUMBIA UNIVERSITY – ‘FIRST-ORDER MAGNETIC

QUANTUM PHASE TRANSITIONS IN MOTT INSULATORS AND

UNCONVENTIONAL SUPERCONDUCTORS.’

Dr. Uemura, Professor of Physics at Columbia University, pointed out the strong connection between

the information revealed by neutron scattering and µSR experiments. In particular, he highlighted past

and ongoing collaborations with neutron scatterers such as P. Dai (Rice), R.J. Birgeneau (UC Berkeley)

and Ch. Pfleiderer (TU Munich) on measuring unconventional superconductors and neutron resonance

phenomenology.

Figure 7. (left) The normalized Bragg peak intensity as a function of pressure in MnSi. (right) The pressure-dependence of the

internal field (frequency) and magnetic volume fraction as measured by µSR.

Figure 6. (left) The asymmetric weak decay of 8Li+ allows this isotope to be used analogously to µ+ to comment on magnetic

properties of materials. (right) Varying the isotope kinetic energy allows a user to choose implantation depth, and student material

properties as a function of distance from a surface or interface.

15

An example of the complementarity of these probes, Dr. Uemura described a collaboration with Ch.

Pfleiderer to measure the quantum phase transition under pressure in MnBi (shown in Fig. 2). While the

magnetic Bragg peak measured by neutron scattering would suggest a 2nd

order transition, µSR reveals

that the since of the internal magnetic field does not change, but rather the magnetic volume fraction goes

continuously to zero at the critical pressure, thereby revealing that the transition is actually 1st order. The

combination of these two techniques, a volume-averaged and a local probe, respectively, reveals more

detailed information about the system than either probe could do individually.

This kind of volume-wise destruction of magnetically-ordered states shows up in many other systems

as well: the Mott insulating rare-earth Nickelates (RNiO3), vanadium oxide (V2O3) and many

unconventional superconductors, including the cuprates and the pnictides. These volume fraction effects

suggest that fluctuations of the ordered and disordered state play a crucial role in determining the nature

of the quantum phase transition. These fluctuations can be probed by by neutrons, for example through

the appearance of resonance modes, and by µSR, often being manifest through an increased relaxation

rate. These techniques will probe fluctuations on different timescales, with neutrons measuring

fluctuations on the order of THz while µSR measures behavior on the timescale of MHz. In combination,

these two measurements can span a much greater range of fluctuations, allowing for a more complete

study of the destruction of the ordered phase.

1.10 JEFF SONIER, SIMON FRASIER UNIVERSITY – ‘SCIENCE WITH µSR: HARD

CONDENSED MATTER.’

Dr. Sonier, Chair of the Department of Physics at Simon Frasier University, began by summarizing a

proposal that was put forward in 2000 to construct a µSR facility at the SNS, for which he was the

primary organizer2. This proposal included a design that used a thin (transmission) target upstream of the

SNS neutron target, which was also discussed during the breakout session (see Section 2.1). While this

design had advantages and disadvantages, the proposal ultimately did not go forward because of a

combined lack of support from the µSR community and from within the facility. Thus, Dr. Sonier

highlighted that the successful construction of a µSR facility being discussed in this workshop would

need more engagement from the community and continued involvement of facility staff at ORNL. He

particularly called on the workshop attendees to make a strong science case for the facility and to remain

identifiably supportive of the project.

The presentation continued on to talk about the increasing complexity of materials and the advances

in devices that use quantum mechanical phenomena. These developments in quantum materials mean

that µSR must be prepared to be able to reliably measure smaller samples under extreme conditions,

including high pressures, low temperature and in high magnetic fields. As an example of these kinds of

developments, a new multi-crystal setup was shown (See Figure 8) that has the capacity for performing

µSR measurements on up to 6 samples simultaneously. The advances in beamline optics and detector

technology would greatly increase the capabilities of a next-generation µSR facility.

16

It was highlighted that µSR is an important probe of these quantum materials, due to its sensitivity to

local magnetism and fluctuations. As an example, the Kondo insulator SmB6 shows evidence for a

topologically-protected metallic surface state3. This has been speculated to occur due to mixed Sm

valence states that undergo Kondo hybridization in the bulk. Measurements on Fe-doped samples of

SmB6 (shown in Figure 9) demonstrate that slowly-fluctuating electronic moments exist in both

compounds and that the fluctuation rates are consistent with Kondo screening. The enhancement of the

fluctuation rate and the smoothing of the transition with Fe doping suggests that the electronic

fluctuations of the Sm moments are the key driving force behind this insulating behaviour. Furthermore,

the μSR measurements are able to demonstrate that there is no phase separation into magnetic and non-

magnetic volumes at any temperature, clearly indicating that this is a bulk property of the material.

Figure 9. (left) Resistivity of SmB6 showing the onset of the bulk insulating state below 10 K. (center) The temperature-

dependent relaxation rate in pure and Fe-doped SmB6. The phase transition in the parent compound is smoothed out in the Fe-

doped case. (right) The μSR spectrum as a function of field, demonstrating the existence of slow magnetic fluctuations.

Figure 8. A picture and schematic of the MULTI sample holder for performing μSR measurements on multiple crystals

simultaneously.

17

1.11 MICHAEL GRAF, BOSTON COLLEGE – ‘MAGNETISM IN α-NAxMNO2.’

Dr. Graf, a Professor of Physics at Boston College, described some of his recent work on α-NaMnO2,

and how it demonstrates the complementary nature of µSR and neutron scattering. The complex nature of

magnetism, through the coupling of various degrees of freedom, means that multiple complementary

techniques are needed. In his example α-NaMnO2, a material that is often studied for its applications to

Na battery storage, there is magnetoelastic coupling. The different aspects of µSR that have been used to

study this compound mirror the different neutron scattering techniques: longitudinal field (LF)- µSR

probes spin dynamics and correlations analogous to measurements of magnetic Bragg peaks and spin

waves, low- and ultra-low-energy muons can probe proximity-induced magnetism in a complementary

way to neutron reflectometry, and even the development of focused muon and neutron beams for their use

in measuring small samples or samples under pressure are very complementary efforts.

Another future direction raised in Dr. Graf’s talk was greater integration of theory and first-principles

calculations alongside µSR experiments. The community would benefit greatly from being able to

routinely and quickly calculate muon stopping sites in the materials being measured, which would

provide more physically significant results. This was highlighted in the data shown, where a calculation

showing two muon stopping sites led to a more reasonable interpretation of the data. Within the neutron

community, there already exists the capability of implementing first principles calculations alongside

experiments; for example, DFT calculations are standard for experiments performed on NOMAD. This

would suggest that a similar approach could be employed at a next-generation muon facility, adding

physical significance to the data collected.

1.12 GRAEME LUKE, MCMASTER UNIVERSITY – ‘MUON SPIN ROTATION/

RELAXATION AS A PROBE OF UNCONVENTIONAL SUPERCONDUCTIVITY.’

Dr. Luke, Professor and Chair of the department of Physics at McMaster University, spoke about the

uses of µSR in probing unconventional superconductivity. Particularly, it was noted that the extreme

sensitivity of µSR to local magnetic fields can be used to measure the field distribution within the vortex

lattice phase of a type-II superconductor. This real-space measurement of the vortex lattice can be used to

measure the magnetic penetration depth, and thereby obtain the superfluid density. A µSR measurement

of the magnetic field distribution in V3Si is shown in Figure 11. The magnetic field distribution in V3Si at

H = 3 T (left) and H = 5 T (right).. At fields just above Hc1, the field distribution is hexagonal, while

closer to Hc2 the field distribution is a square lattice. This change can be seen in the µSR spectra, with

the width and cut-offs of the field distribution being dependent on the magnetic penetration depth and the

coherence length, respectively.

Figure 10. (left) The structure of α-NaMnO2, showing the muon stopping sites. (center0 the triclinic distortions in the monoclinc

bulk structure. (right) The field distribution based on the calculated stopping sites.

18

Figure 11. The magnetic field distribution in V3Si at H = 3 T (left) and H = 5 T (right).

This technique has been shown as a way to distinguish between different types of pairing symmetries

in unconventional superconductors, including the iron pnictide superconductors and iridate

superconductors. Furthermore, the ability of µSR to measure very small magnetic fields can be used to

measure time reversal symmetry breaking (TRSB) fields, in cases of p-wave superconductivity or other

TRSB phenomena such as Sr2RuO4.

Both of these types of measurements have demanding experimental requirements. They often require

low temperatures, applied magnetic fields and very sensitive control of the magnetic field. Thus, a future

muon source would require diverse sample environments and technical support. These measurements

also require high resolution, long counting times and high statistics – requirements that are strongly in

line with a potential µSR facility at the SNS.

1.13 STEPHEN BLUNDELL, OXFORD UNIVERSITY – ‘USING MUONS TO LEARN

ABOUT FE-BASED SUPERCONDUCTORS AND NOVEL MAGNETS.’

Dr. Blundell, Professor of Physics at Oxford University, spoke of his recent experiences exploring

unconventional superconductors and quantum magnets with µSR, placing particular emphasis on

materials containing hydrogenic molecular or organic architectures. Throughout his talk, he emphasized

the complementary role between µSR and other techniques employed in materials discovery efforts. He

noted the continuing positive impact of µSR in the study of new superconducting families. Its relative

speed and sensitivity means that the probe is ideal for constructing phase diagrams detailing the

interaction of magnetic and superconducting order parameters. Because it is a real space probe, µSR is

sensitive to ordering volume fractions, and can address the question of whether phases locally co-exist or

phase separate. Combining these results with x-ray diffraction and other probes, one can also include

structural and electronic information and get a comprehensive picture of the physics guiding new material

families. As an illustration, Dr. Blundell recounted successful efforts at Oxford in recent years to design

and characterize new iron-based superconductor families containing organic spacer molecules. As a first

example, data were presented from a joint x-ray, neutron and µSR study of the newly discovered

superconducting families Ax(NH2)y(NH3)1-yFe2Se2 with A = {Li, Sr, or Rb}, where Tc’s can be controlled

by varying either cation content or (NH2)/(NH3) ratio and can reach as high as 43K (compared with

Tc=9K in non-intercalated FeSe). Time-resolved in-situ x-ray and neutron diffraction (on deuterated

samples) were used to determine local structure and site occupancy, but µSR played a crucial role in

19

demonstrating the existence of a robust superconducting phase in the bulk of the samples, even in the

presence of extensive local disorder. Muon spectra in the vortex phase further allowed for measurements

of the London penetration depths, λ, and showed that these materials followed the famous Uemura

relationship between 1/λ2 and Tc, more commonly associated with the high-Tc cuprate superconductors. A

comparable study was shown for the material Li1-xFex(OH)Fe1-ySe.

As a further example of complementarity, Dr. Blundell discussed his experience studying low-

dimensional magnets, typically materials containing organic molecules and in which interactions between

spins are strongly inhibited (but not eliminated) in one or two dimensions. The enhanced magnetic

fluctuations in these materials, he pointed often obscure signatures of magnetic ordering transitions in

traditional thermodynamic probes such as heat capacity and susceptibility, whereas muons oscillations

remain quite clear. He pointed to several examples from his group in recent years where ordering

transitions were discovered first with µSR, followed by observation of subtle signatures at the ordering

temperature with heat capacity. He then discussed ways in which µSR is helping test theories of low-

dimensional superconductors, and the current efforts at Oxford to “engineer” dimensionality in new

materials through the inclusion of hydrogen ions in existing structures.

During the latter half of the talk, Dr. Blundell discussed the series of "DFT+µ" computer codes

pioneered by his group to help determine favourable muon sites inside materials of interest and quantify

how the presence of the muon charge modify local structure and crystal field environments4. Muon

perturbation effect, he pointed out, thought to be particularly important for systems with delicate ground

states, such as frustrated spin systems, and could be associated with poorly understood reports of

“persistent dynamics” in rare-earth pyrochlores or hamper efforts to detect time-reversal symmetry

breaking fields in non-centrosymmetric superconductors. Specific examples were given to demonstrate

how such computer simulations bolster analysis and aid in understanding µSR data, included ongoing

work on the “quantum spin ice” material Pr2Sn2O7 and spin-liquid candidate α-RuCl3. Determination of

muon stopping sites and perturbation effects was agreed in the room to be a major ongoing concern, and

there was considerable discussion of the computational and theory support desirable at any new muon

source to help address this issue for new users and materials.

1.14 IAIN MCKENZIE, TRIUMF MESON FACILITY – ‘µSR AND β-NMR OF SOFT

MATTER AND CHEMICAL SYSTEMS.’

Dr. McKenzie, Research Scientist and Outreach Coordinator for TRIUMF Meson Facility, discussed

the ways that µSR and β-NMR are impacting the study of chemistry and soft matter science. When

Figure 12. DFT calculations quantify the effect of a µ+ charge on the local crystal field environment in Pr2Sn2O7 (left).

Similar analysis allows for quantitative analysis of µSR data from on a number of 227 pyrochore systems (right). Adapted

from [4].

20

discussing µSR, Dr. McKenzie pointed out that the muon acts as a light isotope of hydrogen, which can

capture electrons to form a metastable “muonium” atom and form chemical bonds with host atoms in the

material under study. He explained how µSR involving such “muonated radicals” can give important

information about local structure, molecular dynamics on ns-µs timescales and reaction rate kinetics. He

further gave showed how muonium can act as a near-perturbationless “spin tag”, much in the same way

nitroxides are used in industry, and garner information about molecular polarity and orientation within

larger structures. Though brief examples were given of experiments performed in time-differential

detection mode, including measurements of proton diffusion and reaction rate kinetics in liquids, the

chemistry examples presented focused on measurements of hyperfine coupling constants (HCCs) of

muons and host atoms, which are measured in time-integrated mode as a function of applied field. These

HCCs give crucial information about electron density, structure of radicals and dynamics. He presented

several examples data sets, including measurements of electron spin-relaxation in organic

semiconductors, the reactivity of biradicals, the orientation of cosurfactants and micelles in bilayers, and

the motion of cholesterolic molecules in a liquid crystal5.

Two interesting examples were also shown to illustrate possibilities using β-NMR as a probe. In the

first, 8Li

+ ions were used as an active probe, and β-NMR spectra garnered measurements of Li-diffusion

rates as a function of temperature in polyethylene oxide, a common electrolyte in batteries. In the second,

β-NMR was performed with 8Li

+ implanted at varying depths in thin-films of polystyrene, giving

evidence for enhanced fluctuations near film surfaces and an associated reduction in observed glass

temperatures with decreasing film size.

Dr. McKenzie pointed out how all of these measurements depended critically on the presence of

appropriate sample environment capabilities- for example the ability to apply several Tesla of field for

measurements of hyperfine coupling constants- and the negative impact of poor beam stability. He

emphasized, however, the central importance of total flux to overall success, especially for time-

integrated measurements. He pointed to the specific comparison of the Hifi spectrometer at ISIS and the

Helios spectrometer at TRIUMF to make his point, where the order-of-magnitude greater flux in the latter

allows for the measurement of more temperatures and applied field points and allows for a more complete

picture of local electronic structure. The talk ended with a discussion about the size of the chemistry

µSR/β-NMR community, which was decided to make up ~10-15% of the user base at European facilities.

It was agreed however that, with the advent of β-NMR and low-energy µSR, there was strong potential

for this community to grow.

Figure 13. Example µSR data on a particular cholesterolic liquid crystal. (left) A schematic of the cholesterol molecule, showing

two muon bonding sites. (center) Time-integrated µSR showing two distinct nuclear resonances. (right) Temperature dependence

of this data reveals the presence of two distinct phases: and isotropic phase (I) at high temperatures, and a “narrowing phase” (N)

where the molecules wobble around a specific axis thus reducing the width of the resonance line. Adapted from [5].

21

1.15 ROSS CARROLL, ARKANSAS STATE UNIVERSITY– ‘µSR & SEMICONDUCTORS.’

Dr. Carroll, Assistant Professor of Physics at Arkansas State University, discussed the role of µSR in

exploring the physics of semiconductors. As with chemistry, muons in these systems act as hydrogen

isotopes, and µSR spectra can give information about the energies, local environment, hyperfine

interaction of interstitial hydrogen sites. This information is invaluable, Dr. Carroll pointed out, for

locating shallow electronic defects and understanding hydrogen passivation behavior in silicon-

germanium alloys, which are used in photodetectors or to make high-mobility heterojunction transistors.

The same information is being used to tailor electronic and optical properties in transparent conducting

oxides, being used in photovoltaic or to make transparent electronics. Also mentioned was a separate line

of research to explore local fields and defect behaviors in a series of magnetic semiconductors, being

explored for spintronics and magneto-optics applications.

Studies of semiconductors typically employ a traditional resonance setup, where nuclear spins in a

strong longitudinal field are perturbed by a transverse rf-field. With the rf-frequency fixed, longitudinal

field is varying, and one searched for resonant loss of muon asymmetry. This is a time-integrated µSR

experiment, which Dr. Carroll stressed depended strongly on total muon flux. As one detailed example,

Dr. Carroll presented his own work searching for shallow defect centers in silicon-germanium alloys.

Using an rf-experiment as described above, his work identified two shallow muonium acceptor levels

with different local environments6. The presence of both sites with different hyperfine interactions and

environments, suggested that both are possible hydrogen defect centers that co-exist in this material.

Dr. Carroll spoke positively of his experiences using neutron scattering to complement his personal

work with muons, and vocally supported strong ties between the two communities. He also strongly

encouraged the development of multimodal capabilities wherein µSR spectrometers are combined with,

for example, the ability to perform in-situ optical pumping. He provided examples showing substantial

modification of relaxation rates of muonium spins in silicon upon illuminating samples with 75W light

from a tungsten source. These observations, along with power and temperature dependences, were used to

extract information about the activation energies and capture cross-sections of muonium sites.

Figure 14. A recent rf-field resonant µSR spectrum from which authors inferred the existence of two shallow muonium acceptor

sites in an alloy of silicon and germanium. From [6].

22

2. SUMMARY OF BREAKOUT SESSIONS

2.1 SOURCE POSSIBILITIES

During the first breakout session, the discussion was aimed at investigating the possible source

designs in more detail in order to analyze their figures of merit and which would be feasible for

construction at the SNS. Three possibilities were considered: a transmission muon target placed upstream

of either the FTS or STS neutron target, a solid muon target utilizing protons being directed to the beam

dump of the accumulator ring, and a possibility that had materialized earlier in the day: a solid muon

target utilizing a portion of the protons obtained by laser stripping the beam from the linac.

The use of a muon target placed upstream of a neutron target has been successfully employed at J-

PARC, ISIS and PSI. Its successful use in other facilities would mean a more straight-forward design and

known source parameters, though it was noted that this method can have a substantial impact on neutron

production. At PSI, for example, ~2% of the proton beam is diverted towards the muon target, but the

associated reduction of proton flux on the neutron target is ~30%, mainly due to beam scatter from the

parasitic muon target upstream. A similar design idea was proposed in 2000 for a potential muon source

at the SNS, utilizing the beam that would be directed to the FTS2. The associated proposal, presented

during the breakout by the primary author, Prof. Jeff Sonier, included 5 µSR beamlines and utilized a 300

ns pulse with a total flux of 1 x 107 µ

+/sec. Alternatively, it was suggested that the resolution could be

increased by chopping the beam to get 10 ns pulses, which would reduce the flux to 6.6 x 104 µ

+/sec. As

the workshop had hit upon an alternate design, presented below, with vastly improved numbers for flux

and resolution, and due to the high potential for impacting neutron production of this design, there was

very little enthusiasm for pursuing this source design further.

A second source location seriously discussed would utilize the protons that would otherwise be

directed to the beam dump of the accumulator ring. In contrast to the design above, this option would

have no impact on neutron production at either FTS or STS. To maximize flux and simplify design, the

suggested location for the muon facility in this design was near the accumulator beam stop on the west

side of the ring rather than near the neutron target buildings- a suggestion that was also noted to be the

preferred option for ORNL , since any additional buildings near the neutron target buildings may impact

future beamline designs or modifications. Furthermore, the placement of an experimental building near

the beam stop had been previously considered for the SEE facility1 and deemed feasible (see Figure 1).

However, significant downsides to the use of beam dump protons were noted, including low beam power

and unfavorable timing structure of the proton pulse. The beam stop absorbs ~2% of the protons from the

linac, with a pulse width of 1ms. To be useful for µSR, a proton source with width comparable to the

pion lifetime (26 ns) is desirable. Any effort to narrow the pulse a reasonable cost would reject 99.9% of

the protons. Overall, attempts to use beam dump neutrons to produce muons would be associated with a

reduction relative to the FTS by a factor of 102 due to power at the beam stop and 10

3 from reducing the

width of the pulse, making it uncompetitive. Again, it was noted that the third design, which used protons

from the linac did not suffer from this significant intensity loss. The longer, 1 ms pulse at the beam stop

would not be unfavourable for a β-NMR source, it was noted, which has a longer decay time. It was

concluded that future discussions may want to consider the feasibility of using the protons directed to the

beam stop for a world-leading β-NMR facility to run in parallel to existing or future materials programs.

By far the most enthusiasm of the breakout session was generated by discussion of the third source

possibility, wherein it was suggested that laser stripping be used to select a proton pulse of controllable

width from the linac prior to the protons entering the accumulator ring, and directing to a muon target

well removed from the neutron production facility. This design presents the opportunity to take advantage

of the entire amplitude of the proton current in the linac with a pulse width that is optimized for µSR. If

23

one used a pulse of comparable width to the pion lifetime, we estimate that one could power a µSR

facility with gain factors of 101-10

3 over existing pulsed µSR facilities (see Section A. 3). Additionally,

the optimized pulse width means that this facility would have the best timing resolution of any pulsed

muon source in the world. The relatively high level of control of laser-stripping technology raised the

well-received prospect of "on-demand" beam timing characteristics, selecting profiles which emphasize

flux or resolution as needed. This capability would be unique to ORNL. As this source option makes use

of protons prior to the accumulator ring, it would have the same benefits as the beam stop design

mentioned above: there would be no impact on neutron production and the muon facility would be well-

separated from the neutron target buildings. Accelerator experts at ORNL confirmed that incident power

can be increased (up to 3%) to entirely compensate for the small (0.15%) proton current directed towards

the muon target. The required laser stripping technology has been tested in limited experiments in the

context for the previous SEE proposal, and shows promise for achieving the functionality necessary to

this design. The principal investigators of this technology are based at ORNL, and so we will perform the

necessary calculations with them to ensure that this design can be successfully achieved.

Finally, it was noted that by utilizing ideas that had been generated during the design of the SEE

facility, such as the laser stripping and building location, it may be an option to construct a single facility

for both purposes. Any future work on a proposed µSR facility should explore the possibility of a joint

SEE test facility, or leave open the possibility of a future expansion for that purpose.

2.2 TARGETS AND BEAMLINES

The second breakout session was aimed at discussing the most effective target and beamline

configurations for the proposed facility. Discussions touched on the merits of having one, two or more

than two muon targets; whether it was feasible to add decay muon and/or low-energy muon beamlines in

addition to conventional surface muon beamlines; and how to plan for future expansion of the facility’s

capabilities.

Focusing on the third source design above with a dedicated proton beam for muon experiments,

discussion centered on dedicated target designs, with the stated aim of maximal use of available protons.

Discussion centered on three main target designs possibilities: a single, solid target; a smaller

transmission target in front of a second, solid target; or a long, solid target that aims to stop the protons

over a range of distances. The first design, utilizing a single, solid target, is the most straightforward and

well understood. Any instruments arranged around this target will receive the same flux of surface

muons, also simplifying the design and versatility of the facility. The second design would feature two

targets: the first would be aimed at stopping a fraction of the muons, with the rest continuing downstream

into a second, stopping target. The benefits of this design are that it would allow a greater production of

surface muons, since each target would be smaller than the previous case, increasing surface to volume

ratio. This concept would also accommodate more room in the experiment hall for more complex

instrument designs. It may also allow the upstream target to operate independently of the downstream

target. However, protons passing through the first target will undergo moderate scattering, reducing the

flux on the second target by ~20-30%. It was noted that this design would also require more shielding

and beamline components, with associated increase in cost. The third target concept considered was a

long extended target, wherein surface muons were produced at the sides of the target over a range of

distances. This novel design would enable an increased production of surface muons relative to the single

target design, while reducing proton scattering that occurs in the second design. However, the production

of decay muons in this design would be diminished, as they are produced in the forward (downstream)

direction in the target. This design concept may also prove difficult to steer the muons into beamlines,

since they are created over a comparatively long length. While no clear consensus emerged from the

discussions of target designs, it was felt that each of these concepts had clear merits that warranted further

24

study. A fourth design presented below, where a kicker element alternately directs protons towards

separate muon targets, was suggested rather late in the discussion and not fully contemplated.

It was realized that the high flux offered by the favored source design would be capable of supporting

beams of surface muons, decay muons, as well as low energy muons. While the surface muons being the

most conventional type of µSR beam, being used for most magnetism, superconductivity and chemistry

experiments, the decay and low energy muon beams would bring substantially more capability to a µSR

facility at the SNS and help expand the community. Decay muon beamlines can provide a means to

enhance the stopping depths, allowing measurements of samples in pressure cells or other containment

vessels, as well as larger, non-uniform components such as mechanical or electronic devices. These

capabilities are an area of strong growth within the neutron program at ORNL, and one or more decay

muon beamlines would bring strongly complementary measurements to this in-house research program.

Similarly, the inclusion of low-energy muon beams would also form strong connections to ongoing

research areas at ORNL. Low-energy beams have a reduced stopping depth, enabling depth-resolved

measurements in thin films, devices and nanostructures, as well as allowing for the measurement of

surface properties. It was noted that enabling a beamline to be switched between conventional surface

muons and low-energy muons would allow for single experiments that can simultaneously measure bulk

and surface properties. Decay and low-energy muon beams are strongly desired by the community; of the

17 current µSR beamlines, 7 are decay muon beamlines and only 1 is a low-energy muon beamline (note:

only 5 of the decay muon beamlines are currently operational; see Appendix C).

Based on the calculations of the muon flux that could be generated at the SNS (see Section A. 2.), it

was felt that 4 surface muon beamlines, as well as 1-2 decay muon beamlines could easily be feasible.

Furthermore, it was deemed scientifically and technically valuable to include the capacity for 1-2

low energy muon beamlines, though specific intensity calculations would need to be performed to

assess feasibility. A world-leading β-NMR source utilizing beam dump protons was also deemed viable,

and could be operated in parallel to the muon source from laser-stripped protons, while sharing building

infrastructure. Such an ion source could be constructed at the same time as the muon facility for minimal

cost, or as a later expansion. More generally, it was felt that the contingency for future expansion should

be considered when creating a full design of the facility. As one possibility, it was noted that it may be

possible to use an electrostatic kicker to direct the proton pulse alternately to two muon targets. This

would require the laser stripping to operate with double the duty cycle, taking a proton pulse every 10 µs,

such that each target would then receive a pulse every 20 µs. In this way, the number of muons and

beamlines at the facility could be doubled if such an expansion was deemed a productive investment. It

would also allow for more capabilities and unforeseen future technologies to be implemented in a wold-

class µSR facility.

2.3 SCIENCE DRIVERS

The ideas presented in the talks stimulated much discussion on the scientific impact of a next-

generation muon source at the SNS during the final breakout session. It was felt that the muon source that

was outlined in the previous breakout sessions would be able to address new and innovative science

questions, owing to the substantial improvement in flux and resolution that this source would allow, as

well as closer integration with computation, detector and sample-environment expertise at ORNL.

In condensed matter physics, the timing structure of pulsed sources create more useful muons per

second relative than continuous sources and offer the opportunity to measure to longer time spectra,

making accessible the detection of smaller moments and slower relaxation phenomena. However, the

maximum moments and faster relaxations, and well as magnitude of applied fields, are tightly constrained

by the timing resolution of the pulsed µSR experiments, tied intimately to intrinsic width of the muon

25

pulse. A muon source at the SNS would have better flux and resolution than any current source of

muons, expanding the range of phenomena available to be measured and while pushing increasing the

flexibility of the technique. In particular, it increases the timescale of fluctuations that can be detected,

which was noted as being strongly complimentary to inelastic neutron scattering. It was also discussed

that the determination of muon stopping sites and perturbative effects, an area of intense current focus in

the field, would strongly benefit from closer integration with computational resources and expertise at

ORNL. Based on the current state-of-the-art at neutron beamlines at ORNL, it was suggested and greatly

encouraged that any new source include specialized software, perhaps on the beamline, which would

greatly increase the information that can be determined from µSR experiments. The complementarity of

µSR to not only neutrons but also traditional thermodynamic probes was noted, and the technique was

demonstrated to be well-suited to the study of phase transitions. This is due to µSR being sensitive to

local fluctuations, real-space phase separation and dynamics on a microsecond timescale, which is in turn

dependent on the resolution and flux of the source. Thus, a next-generation source at the SNS would push

highly sensitive measurements of magnetic phase transitions, already well-developed area of research,

into new territory.

Significant discussion was dedicated to the size of the µSR community, expansion opportunities, and

the positive impact a new source in the United States. One relatively under-developed field, it was noted,

is µSR studies of biological systems. Such measurements been historically limited by the lack of muon

site information, multiple relaxation mechanisms and effects due to muon hopping effects. It is thought

that computational support and increased flux might make a positive impact in this regard. The potential

to include low-energy capable beamlines would also open significant avenues of research in thin films,

devices and nanostructures – µSR research community that has been rapidly expanding since the advent

of low-energy beams at PSI. With the expansion of worldwide research devoted to these areas using all

available techniques, along with recent interest in surfaces and interface effects, low-energy µSR and β-

NMR are expected to be significant growth areas. The workshop participants also noted that the µSR

community has benefitted from several close collaborations with industry in the past few years. In

particular, Toyota has used µSR to study Li diffusion in battery compounds, while Cisco has performed

work studying muon irradiation in electronic components. Industrial partnerships are thought to be likely

to grow in frequency, with various industry partners able to benefit from studying muon irradiation or

hydrogen chemistry through the use of µSR. Such measurements, it is noted, would be greatly enhanced

by the higher flux that would be present at the proposed SNS muon source. Furthermore, by

incorporating new developments in focusing optics, it has been proposed that experiments can be

performed while scanning the muon beam over different regions of the sample. As a local probe with a

small penetration depth, this would provide information on the millimetre length scale across different

portions of the sample. Similarly, focusing optics would enable the measurement of milligram-sized

samples, or samples in pressure cells. The increased muon flux further enhances this capability, allowing

measurements to be performed on samples that are out of the range of current technology.

Through the breakout session, and throughout the meeting, it was repeatedly and strongly expressed

by the participants that the co-location of a muon source with the existing and planned neutron sources at

ORNL would be greatly beneficial for both techniques. The scientific complementarity of the two

experiments has been illustrated above, but it was also highlighted by many that the user communities of

the two techniques also strongly overlap. At two of the European muon sources that are co-located with a

neutron source, ISIS and PSI, 10-15% of the neutron users are also users of the muon source. This is

strongly aided by joint user programs in the two facilities and reciprocal beamtime agreements. With

upwards of 680 unique users at the SNS and HFIR each year, the same rate of users for an SNS muon

source would more than double the number of North American muon source users, without accounting for

the new capabilities made possible by a next-generation muon source design as is being discussed.. It

was also suggested that partnership with the strong programs available for students at ORNL, such as the

26

annual Neutron & X-ray Summer School, would provide another opportunity to increase awareness of the

technique and grow the µSR community in the United States.

27

CONCLUSIONS

As laid out in detail above, the conclusions from this workshop were overwhelmingly positive

with regard to the state of the µSR community and possibilities for a source at the SNS. In addition to

presentations that reinforced the scientific worth of the technique, several strong cases were made that

demonstrate the ongoing development of µSR around the world: increasing user numbers at existing µSR

facilities, increasing collaboration with neutron scatterers and other researchers in materials science, the

opening of the technique to new communities, and several recent discussions for expansion possibilities

for µSR facilities. These developments would only be further accelerated by the construction of a US-

based µSR source.

Indeed, the most exciting conclusion from the meeting was the entirely unanticipated

development of a novel source design at the SNS, which would use existing laser stripping technology

(developed at ORNL) to create a muon source with world-leading numbers for resolution and flux, and a

unique control of beam characteristics. The design was such that the muon source, if constructed, would

have negligible negative impact on the existing or future neutron programs, but a large positive impact on

the overall user program, resulting from the co-location of two powerful probes of matter. The meeting

ended with strong encouragement from all involved that this source design be developed, to both assess

feasibility and further explore possible impacts such a world-leading facility could have on the materials

community in the US and globally.

28

REFERENCES

1 B.W. Riemer and F.X. Gallmeier. Definition of Capabilities Needed for a Single Event Effects Test Facility. Oak

Ridge National Laboratory. Internal Document #ORNL/TM-2014/553. (2014). 2 J.E. Sonier et al. A Next Generation U.S. Muon Facility. (2000).

3 D.J. Kim, J. Xia and Z. Fisk. Nat. Mat. 13, 466 (2014).

4 F.R. Foronda, F. Lang, J.S. Möller, T. Lancaster, A.T. Boothroyd, F.L. Pratt, S R. Giblin, D. Prabhakaran, and S.J.

Blundell. Phys. Rev. Lett. 114, 017602 (2015). 5 I. McKenzie, R. Scheuermann, K. Sedlak and A. Stoykov. J. Phys. Chem. B 115, 9360 (2011).

6 B.R. Carroll. R.L. Lichti, P.W. Mengyan, B.B. Baker, Y.G. Celebi, P.J.C. King, K.H. Chow, and I. Yonenaga.

Appl. Phys. Lett. 105, 122101 (2014).

29

APPENDIX A. SOURCE CALCULATION

30

A. 1. SUMMARY

Following the discussions reported in Section 2, it became clear that the muon source design that was

most feasible and best for addressing the scientific challenges was one that utilized the laser stripping for

directing protons to a dedicated muon target. Compared to a transmission target that was placed upstream

of the FTS or STS neutron target, a dedicated muon target would have no impact on neutron production

or the layout of the FTS or STS target buildings. Furthermore, while a muon target that utilized the

protons that are being directed to the primary and/or secondary beam stop would also have no impact on

neutron proton, but has a pulse timing structure that is not optimal for muon production.

The ability of the laser stripping technique to produce an optimal pulse structure for powering a muon

source, combined with the high power of the SNS accelerator has the potential to facilitate the world’s

highest flux and best resolution pulsed muon source. The following sections outline the calculation of the

muon flux and resolution (Section 30A. 2) of this potential facility, and its comparison to the world’s 4

existing muon facilities (Section A. 3). The workshop was not tasked with preparing a full design, so the

values given below are approximations based on reasonable assumptions about a future facility design.

However, it was the conclusion of all of the workshop participants that these were realistic parameters,

and that this facility would vastly exceed the capabilities of the existing muon sources around the world.

A. 2. DETAILED CALCULATION

After the completion of the PPU project, the accelerator will be operating with the following

specifications:

Accelerator Power 2.47 MW

Proton Energy 1.3 GeV

Current 1.5 mA

Protons per Pulse 2.5 x 1014

Pulse Frequency 60 Hz

Operating Days per year TBD (Currently ~200) Table 1. The operational parameters of the SNS accelerator after the PPU project.

Then the protons per second coming from the Linac can be found by:

Protons per second = (Protons per Pulse) x (Pulse Frequency)

(accelerator) = 2.5 x 1014

x 60 Hz

= 1.5 x 1016

p+/sec (1)

The optimal beam for a muon source is a pulse width that is short compared to the muon lifetime (2.2

µs) and has pulse separation that is long compared to the muon lifetime. However, any proton pulse will

be broadened by the pion lifetime (26 ns) during the proton-pion-muon conversion in the target. Thus,

the optimal proton pulse for this muon source will be 30 ns width, with 20 μs between pul ses. This was

decided through discussion to be the optimal balance of resolution and flux.

Duty cycle = (On time) / (Off time)

= 30 ns / 19.97 µs

= 3 x 10-8

s / 1.997 x 10-6

s

= 0.0015

= 0.15 % (2)

31

The number of protons per second in the muon beam is:

Protons per second = (1) x (2)

(for muon source) = 1.5 x 1016

x 0.0015

= 2.25 x 1013

p+/sec (3)

The laser stripping technique has been demonstrated7 to be ~90% efficient, however at this duty cycle,

the most feasible design would utilize a 4 MW laser in a 12-reflection optical cavity8, which would reduce

the efficiency to ~70% and broaden the time-width of the laser pulse (See Figure 15).

The overall intensity of the laser pulse (left panel of Figure 15) is 50% larger than the optimal square

pulse of 30ns duration, due to the ramping of the laser. Additionally, the full width at half maximum of

the resulting muon pulse (right panel of Figure 15) is 50 ns, containing 77.94% of the total muons

produced. Accounting for both of these factors, as well as the efficiency of the laser stripping, the

number of protons per second that will be diverted to the muon target is:

Protons per second = (3) x (Intensity factor) x (Efficiency)

(to muon target) = 2.25 x 1013

x 1.50 x 0.70

= 2.363 x 1013

p+/sec (4)

The estimate of the proton-pion-muon conversion rate for a solid (thick, stopping) target is difficult to

calculate, since all current targets are transmitting targets, placed upstream of other targets (muon and/or

neutron targets). A Monte Carlo simulation was performed using the specification of the T1 target at

TRIUMF assuming 1.3 GeV incident protons, which gave a conversion factor of 3.6 x 10-7

µ+/p

+ 9, where

we are only counting muons produced from pion decay at rest (surface muons). It is very likely that a

target design that aims to stop all of the incident protons would produce a higher muon flux, but this is

hard to estimate without more in-depth analysis of potential target designs. Using the factor from the

simulation, this would give the number of muons produced, per second, as:

Muons per second = (4) x (Conversion rate)

(from muon target) = 2.363 x 1013

x 3.6 x 10-7

= 8.505 x 106 µ

+/sec (5)

Figure 15. (left) The intensity of the proposed laser pulse as a function of time, with 30 ns at peak intensity. (right) The resulting

time structure of the muon pulse, resulting from the convolution of the laser pulse shape with the pion lifetime.

32

This is the total number of muons produced, which will be distributed evenly over 4π steradians of

solid angle. If we assume that instruments can cover ~ π steradians of solid angle, the number of muons

going to the beamlines will be:

Muons per second = (5) x (Angular coverage) / 4π

(to instruments) = 8.505 x 106 x π / 4π

= 2.126 x 106 µ

+/sec (6)

This angular coverage should be achievable with the use of bending magnets, since the muons are

positively charged, allowing a large fraction of the produced muons to be directed to the beamlines.

If this flux was used to support 4 beamlines, each muon beamline receiving ¼ of the muons produced,

the flux at each beamline would be:

Muons per second = (6) / 4

(per instrument) = 2.126 x 106 / 4

= 5.316 x 105 µ

+/sec (7)

The width of the muon pulse would be a convolution of the width of the proton beam and the pion

lifetime (26 ns). For the shape shown in Figure 15, the width is:

Muon pulse width = [ (Proton pulse width)2 + (Pion Lifetime)

2 ]

1/2

= [ (4.25 x 10-8

)2 + (2.6 x 10

-8)

2 ]

1/2

= 49.8 ns (8)

This is the shorter than any other muon source and offers the distinct advantage that the laser pulse

length would be tunable. So while this calculation offers the flux characteristics for one given resolution,

the entire facility could offer variable resolution depending on the parameters of the experiment.

This is an initial calculation, but would suggest that this source design would lead to a pulsed muon

beam with 50 ns width and 8.5 x 106 µ

+/sec intensity, capable of supporting 4 muon beamlines. It was

also pointed out during the workshop discussion that the proton-pion-muon conversion in the target

would produce approximately 10 times more decay muons (~8 x 107 µ

+/sec), ranging in momentum from

4.119 MeV to 1.3 GeV. These could be used to support one or more decay muon beamlines for

measuring thick samples or samples in pressure cells. Employed as a low-energy muon beamline, the

flux would be reduced by a factor of 104 – 10

5, which would suggest that the muon flux for a low-energy

muon beam would be ~ 4 x 103 µ

+/sec. This suggests that while the resolution would be limited by the

pulse width, the muon flux would be high enough to perform experiments using low-energy muons, being

comparable to the low energy muon flux at the only operational low energy muon beamline. This would

open up new capabilities for measuring thin films and nano-structured materials.

A. 3. COMPARISON TO EXISTING SOURCES

There are four existing muon sources around the world, of which two are continuous sources and two

are pulsed sources (see Table 2). In terms of source characteristics, it makes sense to discuss pulsed and

continuous sources separately. This is because while pulsed sources have much higher flux and can

observe longer signal times, continuous sources have very little background from their lack of pulse

width. This makes pulsed sources more amenable to measuring slowly relaxing phenomena and smaller

magnetic fields, while continuous sources are better suited to faster fluctuations and small volume

fractions.

33

Facility Location Type of Source

ISIS

(Rutherford Appleton Laboratory)

Oxford, United Kingdom Pulsed

J-PARC

(Japan Proton Acceleratory Research

Complex)

Tokai, Japan Pulsed

PSI

(Paul Scherrer Institute)

Villigen, Switzerland Continuous

TRIUMF

(TRIUMF Meson Facility)

Vancouver, Canada Continuous

Table 2. Name, location and facility type of the 4 muon sources around the world.

The flux and resolution (pulse width) are the two most important characteristics for a pulsed muon

source. The source design outlined in the previous section has a pulse width of 50 ns, which gives better

resolution than the source at ISIS (70 ns) and equal to that at J-PARC (50 ns). Furthermore, the high

power of the SNS accelerator, and the ability to strip ~70% of the protons with the laser means that a

muon source at the SNS would have a higher flux (8.505 x 106 µ

+/sec) to either ISIS (1.5 x 10

6 µ

+/sec)

and J-PARC (1.8 x 106 µ

+/sec). The calculation of the flux in Section A. 2 was a first approximation, and

may be increased when designed with a more efficient target. This may result in the proposed facility

having an even higher flux than is currently available anywhere in the world.

34

APPENDIX B. PARTICIPANTS, PRESENTATIONS AND AGENDA

35

APPENDIX B.1 PARTICIPANTS

Table 3. Table of workshop participants.

Name Institution Email address

Adam Aczel ORNL aczelaa@ornl.gov

Stephen Blundell Oxford University stephen.blundell@physics.ox.ac.uk

Ross Carroll Arkansas State University bcarroll@astate.edu

John Galambos ORNL galambosjd@ornl.gov

Michael Graf Boston College michael.graf.bc.edu

Adrian Hillier ISIS, Rutherford-Appleton Laboratories adrian.hillier@stfc.ac.uk

Syd Kreitzman TRIUMF syd@triumf.ca

Graeme Luke McMaster University luke@mcmaster.ca

Mark Lumsden ORNL lumsdenmd@ornl.gov

Gregory MacDougall University of Illinois, Urbana-Champaign gmacdoug@illinois.edu

Andrew MacFarlane University of British Columbia wam@chem.ubc.ca

Doug MacLaughlin University of California, Riverside macl@citrus.ucr.edu

Iain McKenzie TRIUMF iain.mckenzie@triumf.ca

Elvezio Morenzoni Paul Scherrer Institute elvezio.morenzoni@psi.ch

Stephen Nagler ORNL naglerse@ornl.gov

Mike Plum ORNL plumma@ornl.gov

Robert Plunkett Fermilab plunk@fnal.gov

Bernie Riemer ORNL riemerbw@ornl.gov

Koichiro Shimomura Japan Proton Accelerator Research Complex koichiro.shimomura@kek.jp

Jeff Sonier Simon Frasier University jsonier@sfu.ca

Alan Tennant ORNL tennantda@ornl.gov

Yasutomo Uemura Columbia University tomo@lorentz.phys.columbia.edu

Travis Williams ORNL williamstj@ornl.gov

36

APPENDIX B.2 PRESENTATIONS

Table 4. Table of workshop presentations.

Name Institution Title

Stephen Blundell Oxford University Using Muons to Learn about Fe-based

Superconductors and Novel Magnets

Ross Carroll Arkansas State University µSR & Semiconductors

John Galambos ORNL SNS: The Second Target Station (STS)

Upgrade Project Design Status

Michael Graf Boston College Magnetism in α-NaxMnO2

Adrian Hillier ISIS, Rutherford-Appleton

Laboratories

Muons at ISIS

Syd Kreitzman TRIUMF Overview of the TRIUMF Centre for

Molecular and Materials Science

Graeme Luke McMaster University Muon Spin Rotation/Relaxation as a Probe of

Unconventional Superconductivity

Andrew MacFarlane University of British Columbia β-detected NMR at TRIUMF: Status and

Progress

Iain McKenzie TRIUMF μSR and β-NMR of Soft Matter and

Chemical Systems

Elvezio Morenzoni Paul Scherrer Institute Muons for Solid State Research at the Paul

Scherrer Institute

Stephen Nagler ORNL

Mike Plum ORNL SNS Muon Target Possibilities

Koichiro Shimomura Japan Proton Accelerator

Research Complex

Status of J-PARC Muon Science Experiment

Jeff Sonier Simon Frasier University Science with µSR: Hard Condensed Matter

Alan Tennant ORNL Neutron Sciences at ORNL

Yasutomo Uemura Columbia University First-Order Magnetic Quantum Phase

Transitions in Mott Insulators and

Unconventional Superconductors

37

APPENDIX B.3 WORKSHOP AGENDA

Future Muon Source Possibilities at the SNS Workshop

September 1-2, 2016

Thursday, September 1, 2016 Clinch River Cabin

Time Event Speaker

9:00-9:15 AM Opening Remarks and Workshop Goals

Travis Williams

9:15-9:45 am Neutron Science at ORNL Alan Tennant

9:45-10:15 am SNS: The Second Target Station (STS) Upgrade

Project Design Status John Galambos / Mike Plum

10:15-11:00 am Break

11:00-11:30 am Muons at ISIS Adrian Hillier

11:30-12:00 pm Status of J-PARC Muon Science Experiment Koichiro Shimomura

12:00-1:00 pm Lunch

1:00- 1:30 pm Muons for Solid State Research at the Paul Scherrer

Institute Elvezio Morenzoni

1:30-2:00 pm Overview of the TRIUMF Centre for Molecular and

Materials Science Syd Kreitzman

2:00-2:30pm β-detected NMR at TRIUMF: Status and Progress Andrew MacFarlane

2:30-3:30 pm Discussion: Source possibilities, beam profile

3:30 pm- 4:00pm Break / Group Photo

4:00-5:30 pm Discussion: Targets and beamlines

6:00-8:00 pm Dinner “History of ORNL” Steve Nagler

38

Future Muon Source Possibilities at the SNS Workshop

September 1-2, 2016

Friday, September 2, 2016 Clinch River Cabin

Time Event Speaker

9:00-9:30 am First-Order Magnetic Quantum Phase Transitions in Mott Insulators and Unconventional Superconductors Tomo Uemura

9:30-9:50 am Science with µSR: Hard Condensed Matter Jeff Sonier

9:50-10:10 am Magnetism in α-NaxMnO2 Michael Graf

10:10-10:30 am Muon Spin Rotation/Relaxation as a Probe of Unconventional Superconductivity Graeme Luke

10:30-10:50 am Break

10:50-11:20 am Using Muons to Learn about Fe-based Superconductors

and Novel Magnets Stephen Blundell

11:20-11:40 am μSR and β-NMR of soft matter and chemical systems Iain McKenzie

11:40-12:00 pm µSR & Semiconductors Ross Carroll

12:00-1:00 pm Lunch

1:00-2:00 pm Discussion: Science drivers and Future Plans

2:00-2:15 pm Closing Remarks Greg MacDougall

2:30-4:00 pm SNS Tour (Optional)

39

APPENDIX C. COMPARISON OF EXISTING MUON SOURCES

WORLDWIDE

Facility Source Type Flux (µ

+/sec)* Pulse Width (ns) Beamlines

§

ISIS Pulsed 1.5 x 106 70 4 decay

J-PARC Pulsed 1.8 x 106 50

2 surface / 1 decay /

1 low-E

PSI Continuous 7.0 x 104 - 4 surface / 1 decay

TRIUMF Continuous 5.0 x 106 - 3 surface / 1 decay

Proposed at the SNS Pulsed 1.66 x 106 50 4 surface / 1-2 decay

Facility Temperature Range Max Field Max Pressure Co-located neutrons?

ISIS Yes

J-PARC Yes

PSI 0.01 – 1000 K 9.5 T Yes

TRIUMF 0.02 – 900 K 9.0 T 0.5 GPa No

Proposed at the SNS - - - Yes

* Flux represents the time-averaged flux per instrument, averaged across the facility. Note that

continuous sources operating in time-differential mode (measuring one muon at a time) are thus rate-

limited to 7.0 x 104 µ

+/sec.

§ Of the currently-constructed beamlines, the decay muon beamline and one of the surface muon

beamlines at TRIUMF, as well as the decay muon beamline at J-PARC, are not currently operational.

40

APPENDIX D. INVITATION LETTER TO EXTERNAL PARTICIPANTS

Dear Dr. Williams,

Oak Ridge National Laboratory is a world-leading center of neutron science, with excellent

infrastructure for materials research, robust facility user programs, and a broad portfolio of science in

related areas. At the present time a project is underway to develop a conceptual design for a second

target station at the Spallation Neutron Source (SNS). This project would also involve upgrades to the

proton accelerator complex. It is now timely to discuss additional possibilities for enhancing the

effectiveness of the facility, and to this end we wish to carefully consider whether it makes sense

scientifically and technically to utilize protons that would otherwise be directed to a beam dump to power

a SR, NMR or related facility. If feasible such a facility would be complementary to the neutron

science programs and enhance the overall user programs in materials research.

In order to assess the scientific and technical value of such a facility, we are inviting you, along

with other μSR scientists, users and beamline staff, to a workshop at Oak Ridge National Laboratory.

The aim of this workshop will be to discuss the scientific challenges that can be addressed by neutrons

together with a muon / ion source, the technical requirements for building such a facility, and the

perspective of the μSR user community on their desire for such a source. The product of this workshop

will be a Report, which will include recommendations for the feasibility and value of a muon / ion source

and whether provision should be made now for developing such a source in the future. This Report will

be shared with the leadership of Oak Ridge National Laboratory and/or the Department of Energy.

We are currently considering holding this workshop Thursday, August 18th to Friday, August 19

th,

2016, or Thursday, September 1st to Friday, September 2

nd. Please let us know your willingness to

participate and if you are available for either, both or neither of these dates. The dates will be finalized

soon based on participants’ availability. There will be limited funding available to offset at least part of

the travel and accommodation costs for invited participants.

If you have any questions, please feel free to contact either of us and we would be happy to help.

We look forward to hearing from you and hopefully welcoming you to Oak Ridge National Laboratory

for this workshop.

Sincerely,

Travis Williams Gregory MacDougall

Eugene P. Wigner Fellow Assistant Professor

Quantum Condensed Matter Division Department of Physics

Oak Ridge National Laboratory University of Illinois at Urbana-Champaign

williamstj@ornl.gov gmacdoug@illinois.edu

41

REFERENCES

7 Danilov et al. Phys Rev. STAB. 10, 053501 (2007).

8 Y. Liu. Private Communication.

9 S. Kreitzman. Private Communication.